JP6603136B2 - Mapping method for diagnosis and / or prognosis prediction of ulcerative colitis - Google Patents

Description

  A mapping method for diagnosing ulcerative colitis and predicting prognosis by processing the intestinal microbiota in the stool collected from the subject by self-organizing map (SOM) analysis, and such The present invention relates to a method of collecting data for diagnosing ulcerative colitis and predicting prognosis of a subject using an SOM map created by the mapping method.

  Although ulcerative colitis is a disease in which the number of patients is rapidly increasing in Japan, it is regarded as an intractable disease for which no fundamental cure exists yet. In recent years, researchers around the world have attempted to identify the gene responsible for the onset of this disease, but no clear answer has been obtained, including genetic factors, pathological mechanisms, and psychological factors. It is thought that the disease is caused by complicated involvement.

  However, the symptoms of ulcerative colitis have many symptoms in common with other intestinal diseases such as irritable bowel syndrome, including abdominal pain, chronic diarrhea, weight loss and convulsions, and are extremely difficult to diagnose. It has become. Many chronic symptoms, including abdominal pain, chronic diarrhea, weight loss and convulsions, are common with irritable colitis, and the condition may worsen without a definitive diagnosis, while infections such as food poisoning Nevertheless, it is also known that there are cases where ulcerative colitis has been diagnosed and appropriate treatment has not been performed.

In recent years, it has been reported that human health and intestinal flora are closely related (see, for example, Non-Patent Document 1). In basic research using human intestinal specimens, the production of inflammatory cytokines from CD14 ⁺ macrophages, which is the key to the pathological condition, requires bacterial stimulation, and the intensity of stimulation varies depending on the bacterial species. It was proved (for example, see Non-Patent Document 2). In addition, the administration of so-called probiotics, which improves the balance of intestinal bacteria, has attracted attention, and the stool of healthy people is treated and the intestinal bacteria contained in the stool of healthy people are treated with ulcerative colitis The effect is confirmed also about transplanting (for example, refer nonpatent literature 3). In addition, lyophilized VSL3 containing a plurality of bacteria such as lactic acid bacteria, bifidobacteria and streptococci has also been shown to be useful for active ulcerative colitis (for example, non- (See Patent Document 4).

  On the other hand, in recent years, analysis using an artificial neural network (ANN) has been performed in various fields. ANN is a virtual software that performs biological information processing in the human brain, in which about 14 billion neurons (neurons) are connected to each other and perform excellent decision-making activities such as memory and judgment. It is said that this is a system that realizes excellent information processing. Specifically, by learning existing data, it learns by patterning the relationship that exists between the input value and the output value, and also recognizes the new relationship that exists between the input value and the output value. In such a case, the system can predict new data by patterning and learning the relationship (see, for example, Non-Patent Document 5), and is a non-linear relationship that cannot be expressed by a straight line in the natural world. (For example, see Non-Patent Document 6) Analysis is also possible.

Aroniadis OC et al., Curr Opin Gastroenterol. 2013; 29 (1): 79-84. Takayama T, et al., Gastroenterology. 2010 Sep; 139 (3): 882-92, 892.e1-3. van Nood E, Vrieze A, Nieuwdorp M, et al. N Engl J Med 2013. 2013 Jan 16; Online first. Shen J et al, Inflamm Bowel Dis. 2014 Jan; 20 (1): 21-35. Takayama T, et al., Eur J Gastroenterol Hepatol. 2009; 21: 1279-85. Grossi E et al, (2007) Eur J Gastroenterol Hepatol 19: 1046-1054.

  An object of the present invention is to provide means for diagnosis and / or prognosis prediction of ulcerative colitis in a subject.

  In the treatment of ulcerative colitis, administration of useful bacteria to the intestine has been attempted, suggesting the involvement of gut microbiota, while the causative bacteria that are the key to the development of ulcerative colitis are other infectious diseases It is thought that it cannot be specified only in one kind like this, and the bacteria used as a treatment target are not identified. In addition, it was known that the analysis of the bacterial species and the amount of bacteria in the intestinal flora of patients would be essential for elucidating the cause of ulcerative colitis, but data on fecal characteristics and intestinal flora Depending on the conventional statistical analysis used, a tree diagram can be created, but it is difficult to make a decision by associating such a tree diagram with a disease state.

  The inventors of the present invention attempted to examine the verification of the pathological condition and prognosis of ulcerative colitis using ANN. In other words, regarding the prognosis prediction method of patients with ulcerative colitis who underwent cytapheresis, data on patients including at least hospitalization history and surgical history were input factors, and the presence or absence of surgery after cytapheresis was an output factor And applying ANN to the learning; and, when the input factor is input, cytapheresis was performed by operating the ANN using a neuron using a sigmoid function and / or a radial basis function. A method for predicting the prognosis after cytapheresis has already been proposed, including the step of outputting the necessity of surgery in the case (Japanese Patent No. 5484309). This method uses sensitivity (probability of predicting a surgical case to be operated = the number of cases actually predicted to be operated out of cases actually operated / case actually operated) and specificity (for surgery). Probability of predicting a case that did not become surgery = case that did not actually become surgery among cases that did not become surgery / cases that did not actually become surgery) However, when a patient is predicted to undergo surgery, it has not provided a specific means such as reducing the need for surgery.

  This time, the present inventors selected the method of SOM which is one method of ANN, and tried to analyze the gut microbiota of each subject. It was predicted that the relationship between ulcerative colitis patients and enterobacteria was very complex and would not be easily understood, but there were 10 types of enterobacteria in each intestinal flora. By classifying and obtaining analysis data that calculates the total amount of bacterial cells in each classification, and inputting such data as the main input factor and individual background information that is different for each target as a secondary input factor Created a mapping where the cluster was visualized. By using the mapping in which such clusters are displayed, the presence or absence of ulcerative colitis is determined by determining which cluster belongs based on the analysis data of the gut microbiota obtained for the subject. Can determine the pathology of patients with ulcerative colitis, can predict the onset of ulcerative colitis, etc., and if treatment has already been performed, determine whether the treatment method is appropriate, etc. It was confirmed that it was possible to complete the present invention.

The present invention is as specified by the following matters.
(1) The intestinal bacterial flora in the stool collected from the subject is classified into 10 types of intestinal bacterial groups shown in the following table by genetic analysis, and the bacteria for each of the classified intestinal bacterial groups A mapping method for diagnosing ulcerative colitis and / or predicting prognosis by obtaining a sum of body masses and processing data by SOM analysis using the information on the classification and the sum of cell masses as input factors.

(2) The mapping method according to (1) above, wherein the gene analysis method is T-RFLP.
(3) A method for collecting data for diagnosing and / or predicting prognosis of ulcerative colitis in a subject using the SOM map created by the mapping method described in (1) or (2) above.

  By classifying the gut microbiota of each subject and obtaining data on 10 simple classifications, the pathology of patients with ulcerative colitis, the effectiveness of treatment, the frequency of relapses, the onset of ulcerative colitis in healthy individuals Possibility etc. can be determined.

A general dendrogram of the intestinal flora is shown. The SOM map which made the intestinal bacteria data in a healthy subject and an ulcerative colitis patient as an input factor is shown. The SOM map which created the intestinal bacteria data of the patient with ulcerative colitis as an input factor is shown.

  As a mapping method for diagnosis and / or prognosis prediction of ulcerative colitis of the present invention, the intestinal flora in the stool collected from the subject is classified into 10 types of intestinal bacterial groups by genetic analysis, There is no particular limitation as long as it is a method of calculating the total amount of bacterial cells for each classified intestinal bacterial group and processing the data by SOM analysis. Ulcerative colitis mainly affects mucous membranes, erosions and ulcers. Diagnosis of diffuse non-specific inflammation of the large intestine that forms is based on the diagnostic criteria (revised in February 2010) of the Ministry of Health, Labor and Welfare, “Study Group on Intractable Inflammatory Intestinal Disorders” (Suzuki Group).

  As the above gene analysis method, classification into 10 types of intestinal bacterial groups shown in the following [Table 1] in the present invention, and obtaining the total amount of bacterial cells for each of the classified intestinal bacterial groups. There are no particular limitations as long as it is a gene analysis method that can be performed, and a next-generation sequencing method and a T-RFLP (Terminal Restriction Fragment Length Polymorphism) method can be exemplified, but the T-RFLP method is preferable because it is a simpler method.

  In the T-RFLP method, total DNA is extracted from microorganisms in a sample, PCR amplified using a fluorescently labeled primer that recognizes a base sequence encoding a sequence common to all bacteria in 16S rRNA, and the end is fluorescently labeled. Fragment polymorphism that evaluates and compares the number, fluorescence intensity, and position of DNA fragments derived from PCR products of various lengths, including fluorescently labeled ends, after treating the PCR products with restriction enzymes By sex analysis, a method of analyzing the intestinal bacterial community structure can be mentioned, and specifically, a T-RFLP flora analysis method of Techno Suruga Lab Co., Ltd. can be mentioned.

  The classification of the 10 types of enterobacteria groups shown in [Table 1] is used in the T-RFLP flora analysis method, and is classified based on the tree diagram shown in FIG.

  As the “Bifidobacterium” shown in the above [Table 1], Bifidobacterium which is known to act as so-called “good bacteria” such as an immune function enhancing action and an intestinal rot inhibiting action in the intestine is used. Can be mentioned.

  Examples of “Lactobacillales” shown in [Table 1] include so-called lactic acid bacteria including Lactobacillus, Streptococcus, Enterococcus, and the like.

  The “Bacteroides” shown in the above [Table 1] is a normal flora bacterium in the human intestine, but some bacterial species include Bacteroides spp. Known as an opportunistic infection in the clinical field. it can.

  As “Prevotella” shown in the above [Table 1], some bacterial species include Prevotella spp. Known as opportunistic infectious bacteria in the clinical field.

  All of the Clostridial classifications of “Clostridium cluster IV”, “Clostridium subcluster XIVa”, “Clostridium cluster IX”, “Clostridium cluster XI” and “Clostridium cluster XVIII” shown in [Table 1] above are all COLLINS (MD). Based on the report of et al. J. Syst. Bacteriol., 1994, 44, 812-826, it was classified molecularly based on the base sequence of 16S rRNA, and each classification includes Clostridium and / or Includes bacteria other than Clostridium spp. “Clostridium cluster IV” includes Clostridium orbiscindens, Faecalibacterium prausnitzii, and Ruminococcus bromii, which are often detected in human feces. It can be illustrated. “Clostridium subcluster XIVa” reports that many strains that produce butyric acid in the human intestinal tract are contained in Clostridium subcluster XIVa (BARCENILLA (A.) et al. Appl. Environ. Microbiol. 2000, 66, pp.1654- 1661), Clostridium clostridioforme, Clostridium indolis, Eubacterium eligens, Roseburia intestinalis, Luminococcus gnabus ( Ruminococcus gnavus). Examples of “Clostridium cluster IX” include Dialista invisus, Megasphaera elsdenii, and Veillonella ratti, which are often detected in human feces. As "Clostridium cluster XI", there are many cases detected in human feces, and it can be mentioned that it contains bacteria that convert bile acids into secondary bile acids. Clostridium bartlettii (Clostridium bartlettii) ), Clostridium glycolicum. An example of “Clostridium cluster XVIII” is Clostridium cocleatum.

  The “others” is a classification item including a bacterial species that is not included in any of the nine classifications.

  The data processing by the SOM analysis is data processing by an unsupervised machine learning algorithm characterized by competitive learning and neighborhood learning. When data input of an input factor that is a high-dimensional numerical vector is performed, the distance relationship is maintained. This is data processing that learns how to reduce dimensions and convert them to low-dimensional vectors, and provides assistance for visually understanding the analysis results of high-dimensional data.

  As an apparatus for performing data processing by the SOM analysis, there are initialization processing means, input means for setting information of an input vector consisting of a plurality of attributes, and the same number of input vectors as most similar to the input vector. Means for searching for a winner unit having a reference vector made up of attributes, means for updating each unit based on a learning coefficient that is sequentially reduced according to the number of learning and the input vector, and two-dimensional display of the unit An apparatus having means for creating a map to be created can be mentioned.

  Examples of the initialization processing means include means for setting a provisional arrangement position of each unit in the map before inputting individual data. The vector of each unit is set at random using a random number. It is preferable. By this initialization process, the learned process is appropriately executed, the failure of learning can be prevented, and the integrated map can be converged with higher accuracy and higher speed.

As an input means for setting the information of the input vector consisting of the plurality of attributes, X1 (x, which is data relating to the sum of the bacterial mass for each of the 10 types of enterobacteria groups and the classified enterobacteria groups ₁ , x ₂ , x _3, x ₄ , x ₅ , x _6, x _7, x ₈ , x ₉ , x ₁₀ ), X 2, X 3, X 4. A means for inputting an input factor can be exemplified. In addition, when inputting data for each of the above main input factors, whether n people are ulcerative colitis patients, whether they are healthy (non-ulcerative colitis patients), family history data, Data on ease of relapse, activity of ulcerative colitis, data on the extent of ulcerative colitis, data on the duration of ulcerative colitis, data on clinical types of ulcerative colitis, intractable Data related to sex (steroid resistance or steroid dependence), data related to extra-intestinal complications, data related to abdominal symptoms (frequency of stool, stool characteristics, presence or absence of bloody stool), blood sampling data (red blood cell count, hemoglobin level, white blood cell count, CRP level, Individual protein basics such as total protein level, albumin level), endoscopic findings (endoscopic sub-score of Mayo score), past illness, gender, height, weight, age, etc. Data can be entered as a secondary input factor.

Examples of means for searching for the winner unit include means for searching for a winner unit U _win (w ₁ , w ₂ , w ₃ ... W ₁₀ ) closest to Xn.

As a means for updating each unit (U), first, the value of U _win is made closer to X1 according to U _new = U + α (X1−U _win ) (where α is a learning coefficient), and further U _win Although the value of the unit in the vicinity of X is also brought close to X1, there can be mentioned means for changing the value of each U by decreasing the rate of approaching X1 as the distance from U _win increases. X2, X3, X4,... Xn are also sequentially input to perform learning repeatedly, and the learning coefficient is sequentially reduced according to the number of learning. As a result, data having a high degree of similarity gathers nearby, data having a low degree of similarity is mapped far away, and a two-dimensional SOM map in which a plurality of clusters are formed is created.

  In the mapping method of the present invention, the shape of the unit constituting the map to be created is not particularly limited. For example, a quadrangle or a hexagon can be mentioned. If the unit shape is a quadrangle, the nearest unit is four. In the case of a hexagon, the closest unit is six.

  As the information of the input vector in the present invention, the types of intestinal bacteria (group) obtained by classifying the intestinal bacterial flora of each target into 10 types by the above-described genetic analysis method, and the bacteria of each intestinal bacteria (group) Data on the sum can be listed. In addition, although not used for classification of each cluster, by inputting the stored data as a secondary input factor on the helix where the data is stored, the analysis of the secondary data in each cluster group is combined. Can also be done.

  Examples of the secondary input data include data on items related to clinical information of each subject (human), specifically, whether the subject is a patient with ulcerative colitis or a non-ulcerative colitis. Data, family history data, data on the presence or absence of relapse, basic data on items judged by health checkups such as individual gender, height, weight, blood pressure, etc. It is preferable to input two secondary input data, and the items of the target secondary input factor are preferably matched.

  The subject in the present invention may be an ulcerative colon patient or a non-ulcerative colitis patient. As the subject is an ulcerative colitis patient, the Ministry of Health, Labor and Welfare “Survey Research Group on Intractable Inflammatory Intestinal Disorder” According to (Suzuki group) diagnostic criteria (revised February 2010), a) clinical symptoms, b) endoscopy or enema X-ray examination, c) biopsy histological examination meets diagnostic criteria, By excluding the disease, a person diagnosed with ulcerative colitis can be mentioned.

  When the subject is a non-ulcerative colitis patient, if the subject is a healthy person, suffers from a disease other than bowel disease, is not ulcerative colitis, enteric Behcet's disease, Crohn's disease, etc. Cases with other bowel diseases can be included.

  By using the map created in the mapping method of the present invention, a subject who has received a determination for the diagnosis and / or prognosis prediction of ulcerative colitis, the subject's own data is input, The map included as a target in the mapping method of the invention and created by the mapping method of the present invention is more accurate, and data for diagnosing ulcerative colitis and / or predicting prognosis of the subject Can be collected.

  Examples of the data on the family history include the presence or absence of a history of ulcerative colitis up to the second generation (grandparents), that is, a bloodline within the third degree.

  Examples of the data relating to the ease of relapse may include data relating to the relapse rate of patients determined to be in remission (Lichtiger index is 4 or less) for patients undergoing treatment for ulcerative colitis. Although the relapse rate is about 30 to 60%, which differs depending on the report, this time, a case that relapsed twice or more a year was defined as an easy relapse case.

  Basic data of the above individuals include the presence or absence of ulcerative colitis, past history, family history such as cancer, drinking history, smoking history, gender, height, weight, age, defecation status (defecation frequency, stool characteristics), The intake frequency of yogurt, the presence or absence of hay fever, etc. can be mentioned.

  Data on the effectiveness of the therapeutic effects of the above-mentioned drugs include oral and rectal administration, which suppresses persistent inflammation, is effective for mild to moderate ulcerative colitis, and is effective in preventing relapse. It is effective for the administration of 5-aminosalicylic acid (5-ASA) such as salazosulfapyridine (sarazopyrine) and mesalazine (pentasa and asacol), and has the effect of strongly suppressing inflammation, and is moderate to severe ulcerative colitis Of corticosteroids such as prednisolone (predonin) and steroids that are effective for the treatment, but do not have a relapse prevention effect, or steroids that show a certain effect but worsen with weight loss. Azathioprine (AZA), 6-mercaptopurine (6-MP) that are effective for the subject; Recorded effectiveness / ineffectiveness of administration of immunosuppressive agents such as cyclosporine and tacrolimus, which are effective, and administration of anti-TNFα receptor antagonists such as infliximab (remicade) administered by infusion and adalimumab administered subcutaneously For example, effectiveness data.

  The data on the effectiveness / ineffectiveness of the above-mentioned drug treatment include blood cell component removal therapies such as LCAP (LeukoCytAPheresis) and GCAP (GranuloCytAPheresis) and stool of healthy people in patients with ulcerative colitis Data on the efficacy of fecal transplantation, VSL3 and other probiotics, fermented foods, oriental medicine such as acupuncture and herbal medicine, and antibiotics Can be included.

  According to the mapping method of the present invention, it is possible to determine the presence / absence of ulcerative colitis in the subject, the pathology of the ulcerative colitis patient, the effectiveness of the treatment being performed, the frequency of relapse, and the like.

  EXAMPLES Hereinafter, the present invention will be described more specifically with reference to examples, but the technical scope of the present invention is not limited to these examples.

[Example 1]
[Sample collection]
Under sufficient informed consent, a stool specimen of about one azuki bean from 100 healthy subjects and 100 patients identified as ulcerative colitis was stored in a highly airtight preservation solution (guanidine solution). The sample was collected in a container and immediately stored in a refrigerator. The sample was then refrigerated and transported to Techno Suruga Lab Co., Ltd. for analysis by the T-RFLP flora method. Based on the above [Table 1] of the intestinal bacteria of each subject analyzed by the T-RFLP method, data on the classification of the intestinal bacteria group and the sum of the bacterial mass of each intestinal bacteria group were obtained.

[Create SOM map]
SOM initialized for each of the subjects of 20 healthy subjects and 100 ulcerative colitis patients, using as a main input factor the data on the classification of the acquired enterobacteria group and the total amount of cells for each enterobacteria group Input to analysis software. In addition, when inputting each data, whether the patient is an ulcerative colitis patient or a healthy person (non-ulcerative colitis patient) was input as a secondary input factor.

  A two-dimensional SOM map with 5 clusters was created as shown in FIG. Analysis of secondary data for these five clusters revealed the characteristics shown in [Table 2] below.

  Using the SOM map prepared above, analysis was performed for an unspecified subject. The fecal intestinal flora collected from each subject is analyzed by the T-RFLP flora method, classified into 10 types of intestinal bacterial groups, and the bacterial cells for each of the classified intestinal bacterial groups Data on the total amount was obtained. After determining which cluster of [Table 2] such data belongs, it can be determined as follows.

  (1) In cluster 1 (light blue) and cluster 3 (yellow), ulcerative colitis patients and healthy people were mixed. Therefore, when the subject classified into the cluster 1 or 3 is an ulcerative colitis patient, the pathological condition is considered to be improved, while when the subject is a healthy subject, the ulcerative large intestine is considered. It is thought that there is a possibility that the flame will take in the future. For cluster 1 and cluster 3, women were often classified as 1, and 3 tended to have more men. For patients with ulcerative colitis, cluster 1 tended to be more symptomatic than 3.

  (2) In cluster 2 (red), most were patients with ulcerative colitis. Therefore, when the subject classified into cluster 2 is a patient with ulcerative colitis, it is considered that the disease state has not improved so much and the administered drug is likely to be ineffective. In particular, if a patient who belonged to cluster 1 or 3 in the previous analysis now belongs to cluster 2, the pathological condition has deteriorated and it is considered that the treatment method should be changed. However, if a patient who belonged to cluster 5 in the previous analysis now belongs to cluster 2, it can be considered that the pathological condition has improved somewhat. If the subject is healthy, it is likely that they will develop ulcerative colitis in the future

  (3) In cluster 5 (purple), all patients were ulcerative colitis. Therefore, when the subject classified into the cluster 5 is a patient with ulcerative colitis, it is considered that the pathological condition has not improved so much and the administered drug is likely to be ineffective. In particular, when a patient who belongs to another cluster in the previous analysis is to belong to cluster 5 this time, it is considered that the treatment method should be changed. If the subject is a healthy person, it is very likely that they will develop ulcerative colitis in the future, and even if there is no subjective symptom, there may be inflammation in the large intestine. It is done.

  (4) In cluster 4 (green), most were healthy individuals. Therefore, when the subject classified into the cluster 4 is an ulcerative colitis patient, it can be determined that the administered drug is effective and the disease state is improving. If the symptom is severe, it may be due to other bowel disease rather than ulcerative colitis. In addition, if the subject is a healthy subject, it is considered that the possibility of suffering from ulcerative colitis in the future is low.

  The subject who has acquired data on the classification of the intestinal microflora can be determined as to which cluster to classify according to FIG. 2, and can be judged on the pathology, etc. The above SOM map is learned by inputting as, and a secondary input factor is input as to whether the patient is an ulcerative colitis patient or a healthy person (non-ulcerative colitis patient). A high SOM map can be created.

[Example 2]
[Sample collection]
With sufficient informed consent, a stool sample of about 1 azuki bean sample from 200 patients identified as ulcerative colitis is quickly collected in a highly airtight container containing a preservation solution (guanidine solution). After refrigerated storage, the sample was refrigerated and transported to Techno Suruga Lab Co., Ltd. for analysis by the T-RFLP flora method. Data on the classification of the intestinal bacterial groups of the intestinal bacteria of each subject analyzed by the T-RFLP method based on the above Table 1 and the total amount of the bacterial mass of each intestinal bacterial group was obtained.

  For each subject of 200 patients with ulcerative colitis, data on the classification of the intestinal bacterial group and the total amount of bacterial cells in each intestinal bacterial group were input as (primary) input factors into the software for SOM analysis, Data on family history, data on ease of relapse, presence of activity, data on disease extent, data on disease duration, data on clinical disease type, data on intractability (steroid resistance or steroid dependence), extra-intestinal complications Data, abdominal symptoms (frequency of defecation, stool properties, presence or absence of bloody stool), blood collection data (red blood cell count, hemoglobin level, white blood cell count, CRP level, total protein level, albumin level), endoscopic findings (Mayo score) Endoscopic sub-scores), past history, sex, height, weight, age, etc., individual basic data of the target individuals were entered as secondary input factors.

  A two-dimensional SOM map with four clusters was created as shown in FIG. Analysis of secondary data for these four clusters revealed the characteristics shown in [Table 3] below.

  As is clear from the above [Table 3], the SOM analysis also revealed a cluster having a correlation with the relapse frequency and family history.

  By using the map shown in FIG. 3 based on the characteristics shown in [Table 3] above, it is possible to determine the following for unspecified new subjects in the treatment of ulcerative colitis and the like.

  (1) Various conditions were mixed in cluster 1 '(light blue). Therefore, if the subject classified as cluster 1 ′ is a patient with ulcerative colitis, the condition is considered to be improved, while if the subject is a healthy person, it will suffer from ulcerative colitis in the future. There seems to be a possibility.

  (2) In cluster 2 '(red), many patients relapsed and many had a family history. Therefore, if the subject classified as cluster 2 ′ is a patient with ulcerative colitis, there is a high possibility of relapse even if the subject is in remission, and if the subject is healthy, the subject will develop ulcerative colitis. This is likely to happen in the future.

  (3) In cluster 3 '(yellow), the degree of inflammation was strong, and many patients were in poor condition. Therefore, if the subject classified as cluster 3 ′ is a patient with ulcerative colitis, the pathological condition is considered to be very bad, and if the subject is a healthy person, ulcerative colitis may occur in the future. It is considered that there is a possibility that the large intestine is inflamed even in the absence of subjective symptoms.

  (4) In cluster 4 '(green), there were few subjects who relapsed. The subjects were also relatively older men. Therefore, for example, if a subject who belonged to cluster 3 ′ in the previous analysis now belongs to cluster 4 ′, the treatment method is appropriate, the pathological condition is ameliorated, and there is a possibility of relapse. Is considered low. In addition, when a new subject has an intestinal microflora pattern similar to that of a subject belonging to the cluster 3 ′, a treatment method similar to that for the subject belonging to the cluster 3 ′ is used. It can be assumed that it is useful. If the subject's symptoms are not alleviated, it may be due to other bowel disease rather than ulcerative colitis. In addition, if the subject is a healthy subject, it is considered that the possibility of suffering from ulcerative colitis in the future is low.

(Summary)
The clustering obtained by the mapping method of the present invention is not based on the known classification method of the gut microbiota pattern in patients with ulcerative colitis, but is a technique for discovering a new useful classification method. It is a period. Unlike simple classification, when a patient with a new ulcerative colitis is the subject, it helps to improve the disease state by judging the treatment method from the correlation with the disease state, the effectiveness of treatment, and the frequency of relapse. It becomes.

  The present invention is very useful in the medical field.

Claims (3)

The intestinal bacterial flora in the stool collected from the subject is classified into 10 types of intestinal bacterial groups shown in the following table by genetic analysis, and the amount of bacterial cells for each of the classified intestinal bacterial groups is classified. A mapping method for diagnosing ulcerative colitis and / or predicting prognosis by obtaining a sum and processing data by SOM analysis using information on the sum of the classification and the amount of cells as an input factor.
The mapping method according to claim 1, wherein the gene analysis method is T-RFLP. A method for collecting data for diagnosing and / or predicting prognosis of ulcerative colitis in a subject using the SOM map created by the mapping method according to claim 1.